KGB- DC Coupled Hybrid Balance OTL/OCL HeadAmp

[aparatusonitus]

Hi guys,

I belive this is my first post at tube section, 'cose I've been busy with sand, sorry. Recently I manage to buy 16 Ei PL509 along with Magnoval sockets wery cheaply. I also have a Sennheizer HD600 (300 Ohm) and Philips SBC-HP890 (32 Ohm) headphones but not a tube headamp capable to drive them both. So I am in the mood for something really special, like DC coupled balance OTL headamp based on PL509. I found only one schematic that satisfied my needs, that is schematic by dr. Kevin Gilmore (of head-fi.org fame) OTL based on 6C33C-B. Original sch have some mosfets which I can't get, so I changed that part of circuit for IRF9231 and IRF610 along with output tubes, some resistors and pots, added snoobers network and inductive decoupling at output. I think that servo opamps is a must in amplifier like this, but I have not try to solve that part yet. Guess that each output (+ & -) will need it's own servo opamp to control relative DC offset (+ output agaist ground and - output against ground) & one more to control absolute DC offset (+ output against - output). Let me tell you that I am NOT an EE nor have long experiece with SS and tubes, so ANY help and advise is wellcome.
And yes, I am willing to split ground in my headphones if this Godzilla works

Best, a

[aparatusonitus]

FFT at 1Vp-p output...

[aparatusonitus]

FFt at full tilt 19v p-p

[aparatusonitus]

Gain, not needeed that much, but anyway...

[aparatusonitus]

AC analysis...

[aparatusonitus]

Transient at full tilt...

[kevin gilmore]

Wow. I wanna see pictures. Now replace the current
sources with tubes and you have a MIG2.

I'll analyze all the currents and voltages and see if i can't
come up with a few minor changes. It should be stable enough
not to need a servo...

[aparatusonitus]

Hi Kevin,

Thank you for replay, I really need all the help I can get from gurues around here.

Now, as suggested, I managed somehow to swop BJT current source for tube one, but I have to learn that part first from Bruce Rozenblit book. How does it looks? At this point I should point out that, if this Headzilla ever come to life, I intend to bild poweramp base on same (more or less) sch. Therefore, I have to spare some of the PL509 for that duty, and for that reason I would rather use SS (M-fet or BJT) current source no matter how tube CCS looks good on spice or actual sound/measurement of a real thing.

That appart, I think I have trouble with Vbe mutliplier (Q3/Q4)and M-fet differential amp (M2/M3). Should I swop BJT Vbe multipier for M-fet Vgs mutipier (something like ZVN3306) and keep the 2 x IRF610 diff. amp with some resistors/pots values changes?

Any way, here is KGB v1.1 with tube current source...suggestions, comments?

Regards, a

[aparatusonitus]

Servo and some compensation added... Could someone take a look, please

[Tubes4e4]

Hi aparatusonitus,

I wonder what the aim of this exercise might be - just to show that such a Rube Goldberg monstrosity could be done somehow?

BTW, the EL509/PL509 is a horrible tube to start with, even in trioded mode. See this link for the plate curves., moreover, it is really hard to keep from drifting.

A decent OTL headphone amp probably could be done with about 5% of the parts count you are at already in your schematic, even if you decide to go hybrid, which usually tends to up the parts count considerably, compared to pure tube tech.

Just my 2 Euro-cents.

Tom